What Clarecast Data Reveals About AI and Quiet Restructuring

[00:00:00] The leading issue of agentic AI in businesses right now is ensuring agents act with compliance guidelines. And Denodo applies guardrails across your entire data estate. By aligning your company's data infrastructure under one system, these guardrails perform consistently across your platform. So start scaling your business and start with Denodo. Simply visit denodo.com to learn more.

[00:00:27] If your news feed is to be believed, AI has already taken everybody's job. It's rehired half of us as a chief AI officer and scheduled another restructuring before lunch. The data, however, tells a much more complicated story.

[00:00:49] That's one of the reasons I'm excited to be joined by the chief data scientist and co-founder of a company called Clarecast on the podcast today. Because the company analyzes employment profiles, hiring activity, technology adoption, company information and industry signals to forecast how businesses may develop over the next 18 months.

[00:01:13] So my guest today will discuss their research into quiet restructuring while companies may slow hiring, leave positions unfilled, but also introduce new ones or reorganize internally, all without announcing mass layoffs. And he also will explain why AI-related announcements does not automatically prove a change in headcount, despite what you might read on your news feeds.

[00:01:41] So today we're going to hear how to distinguish probability from prediction, why workforce data always carries a delay, and how leaders can use predictive intelligence without allowing a confident model to turn uncertainty into certainty. This is a timely reminder that headlines move faster than evidence. So we're going to put all the emotion to one side today and see what the data is telling us.

[00:02:10] But enough from me. Let me introduce you to my guest now. So thank you for joining me on the podcast today. Can you tell everyone listening a little about who you are and what you do? Hi. Thanks, first of all, Neil, to invite me to this podcast. I really appreciate you giving me this opportunity. My name is Marvin. I'm, as we just talked in the pre-show here, I'm sitting in Michigan, Grand Rapids. You might wonder. This is a woman boy sitting in Michigan, Grand Rapids.

[00:02:42] So I've been in a lot of places. So I originally studied physics actually in Germany and Berlin after I finished my PhD. Actually during my PhD, I got to know a lot of interesting people. One of those was a professor from Berkeley. So that gave me a chance to, you know, after I finished my PhD in Germany to move over to the US, spend some time in the San Francisco Bay Area at the Berkeley Lab. So it was kind of a shared responsibility between Berkeley Lab and UC Berkeley.

[00:03:12] My work at the time was, you know, I was working in a physics lab. We were doing physical chemistry. It was a lot of, you know, spectroscopy work. It was literally like, it was a very holistic work. When I'm looking back at it, it was very fun. So there were aspects, there were days where I'm just, you know, sitting under the vacuum pump. I was, you know, fixing screws or I was soldering something. So really hands-on work. There were days where, you know, you have to write in proposal, write in paper, where it's really like sitting at your desk all the day.

[00:03:41] And of course, all that, you know, quantitative aspects around it. Like we were measuring data, we were analyzing data, we were publishing data. So I kind of had a broad spectrum of, you know, experiences I had during this time, which was really fun. But at some point I was looking around and thinking for myself, okay, where do I go from here? What's going to be the next step for me? What I'm actually interested in? And also, you know, what I'm interested in, but also what gives me a certain ROI, right? So I want to also, you know, go into a career that's fruitful for me.

[00:04:12] And I must say my time during Berkeley, this was really transformative in a way, because it was between 2019 and 2022. And it was before the big inflection point of OpenAI, ChatGPT. So at that point, you couldn't go on the street and, you know, ask a random person, have you ever heard of AI? They would say, sure, AI, that's the thing in TV, right? That's these robots that talk to you, right? This is what AI was at the time.

[00:04:39] But back in the day in Berkeley, like the Y-Ape was totally different. Like you moved around campus and, you know, everyone was already talking about it. Like everyone had a side project, you know, trying something out. Not even in STEM, like even if you're in the social sciences, you went at these, you know, facilities and people say, hey, this is my study. And try to kind of, you know, mix AI into it. I tried to do machine learning. And machine learning was basically the proxy route for AI back in the day.

[00:05:08] And this is kind of the experience that I had where I thought, okay, I'm going to look into it. And of course, during my time as a research scholar, I mean, we did a lot of scripting, right? We had to analyze our data and so on. But I wasn't really a coder or software developer. So even though it's not 100% true, I always like to say, okay, I didn't write a single line of code until 2019. It's not 100% true, but it's close to it, right? And today we have 2026.

[00:05:38] And now I'm the, you know, chief data scientist and co-founder at an AI startup. So this is kind of how my life progressed from there on. Wow. What an incredible journey you've been on there. And I love how you've gone from academic physics and research at Berkeley to building AI systems and now predictive intelligence at ClaireCast.

[00:05:59] So I've got to ask, how has this journey shaped the way you think about what separates an interesting AI model from one that can be trusted to inform real business decisions? Because you've been at every stage here and working in AI before it was cool and everyone was talking about it. You've been there from the beginning. How has all this shaped how you approach AI? Yeah, yeah, yeah. That's a very interesting question.

[00:06:23] And I must say, even though I'm, you know, in the grand stream of things for a very short period in this whole industry, I've seen so many mental shifts from, you know, people, business leaders, how they see AI. So going back to my journey, after I decided, okay, I want to do a career in tech, I decided to become a data scientist. I acquired all the, you know, skills and all the weaponry to go to this industry.

[00:06:53] So I went back to Germany. It was in late 2022 and worked at a chemical, you know, a large chemical company, BASF in Germany as a data scientist. So back in the day, AI, data science, it was kind of, you know, if people talk about data scientists, they really meant AI. When they talked about AI, they really meant like it's probably the data scientists that work on it. But data science in its clearest form, in its rawest form is, you know, it's a lot of statistics.

[00:07:23] You're doing, you know, quantitative analysis. You're doing machine learning. So this is kind of the, you know, the work you do as a data scientist. But then, you know, as I already alluded, we had 2022, JGBT came and this is when like, now we have large language models. We call them foundational models because they can do anything, right? You can do, you know, a classification with them. You can do, you know, next token prediction with them. You can, you know, these days even create, you know, whole Hollywood movies or we are on the brink of actually doing this.

[00:07:51] So these days it's totally different. Right. And also when I look at back at 2022, 2023, so I kind of was the guy, okay, you're the guy you're working with AI. Can you please also be the guy that's transforming our business? I mean, this was the big, the big topic in businesses, which is still today. Like, okay, we have this organization and we kind of need to educate our people, right? We need to transform our processes. And you're not just, you know, by a new tool and then your process is transformed.

[00:08:20] You have to, you know, take your, you know, you have to take your organization. You have to take people in the organization and show them and educate them, you know, what they're going to do with these new technologies. So, and I remember back in the day when I had these, you know, tone halls and meetings where we were explaining AI, what you can do. So people were always like super, they were reserved in a way.

[00:08:44] They were like, they were coming to me and saying, hey, I tried, I tried out this new LLM tool and I asked it some random question about my expertise. And the answer it gave was just totally wrong. So, so this was back in the day. My job was basically to push back, back in the day to say, hey, that's sure. That's what we call, that's, that's, that's made up. That's what we call hallucination. It's because the model doesn't have this information and, you know, training set. So it just gives you some information very confidently, but it's not right.

[00:09:12] And people were, were always trained in like, okay, they put a question to Google or any other search engine and they were getting the right answer. This is not how these AI models work, right? They give you any answer. So they're kind of language machines. There's a different way how to work with them. So this was back in the day, but now fast forward, like today and maybe last year, like my role in this organization totally switched. Now I was the guy, hey, by the way, these models, it's still kind of dangerous. You shouldn't do anything with them. You shouldn't believe everything they say.

[00:09:41] They try to be super confident and they try to be sycophants. They try to, you know, always confirm your biases. Please be cautious. So now I'm on the other side of the coin, right? People always say, hey, my AI told me this is how we should lead the business. And the AI says that, so should we follow it? Now I'm saying, no, no, no, step back. It's still a computer somewhere in a data center that's trained to, you know, give you an answer that you like. So it's, you know, this whole, like the reinforcement learning works, right? It's designed to tell you what you want to hear.

[00:10:12] So, yeah, that's kind of a funny perspective switch that I observed within a short amount of only two years. Such a great point there. And for people listening that are hearing about ClaireCast for the very first time, can you tell them a little about what you're building and how you're actually turning signals across millions of companies, employment profiles, job postings and tech adoptions into predictions about where a business might be heading without the sycophants there as well.

[00:10:40] So tell me more about that and what you're building. Absolutely. So I came together with my two co-founders. One co-founder is Jonathan. I met him during my time at Berkeley. He was working at Databricks at the time, which is, you know, one of the heavy hitters when it comes to data and AI platforms. We were in different paths, but always kept in close context.

[00:11:01] And he reached out to me sometime last year telling me, hey, I have this idea and I have met this person, which is the third co-founder, Bradley. And we're thinking about starting a startup and we want to talk to you. And this was the first time when, you know, we as a group, as co-founders come together and, you know, sparking those ideas and bouncing those ideas across each other. And Bradley, he's coming. He's who's the CEO of our company.

[00:11:30] He has this background in, you know, he built a company in the insurance space and the employee benefit space. He exited. Then he was working at a, you know, PE portfolio company for a while in a strategy officer, chief strategy officer position. And he always had the struggle when he wanted to, you know, predict and look forward, like how his own business is developing.

[00:11:56] He had to know, okay, where the market is going because he was at the business insurance, right? Your business, your revenue is basically a proxy of how your customer base is growing or shrinking. Because an insurance, it's like, it's inherently seed based, right? If a company is adding headcount to its company, they also need insurance for these practices. If they are shrinking, also like the insurance carrier is also shrinking because they can sell less policies.

[00:12:24] And this is, of course, true for insurance, but it's also true for almost every, you know, SaaS startup of the last decade, right? It's always licensed. It's seed based, right? If you add a new person, you need a new Salesforce account. You need a looks like an account or whatever tool you have in your company. So all these SaaS startups, they also base their revenue on headcount.

[00:12:46] So the big problem is if you want to plan your business, you kind of, you know, in a way you can, you know, you can try to foreshadow the future by looking at the past. But that's a very reactive way to, you know, plan your own business. So we came together and said, okay, there might be a more clever way, right? We have all these tools. We have AI. We have machine learning.

[00:13:09] There are like these trillions of data sources out there that you can tap into and try to get a little bit more sophisticated, a little bit more, you know, a little bit more educated in how you make these forecasts. And this is where we came together. We said, okay, we want to be a predictive business intelligence platform, right? It's kind of a lot of words, but what we basically want to be is, I mean, the important word here is the platform, right?

[00:13:36] The idea is that we want to basically provide the fundament to build a lot of use cases off of it. One of those use cases is you might just, you know, want to get your portfolio companies into our platform. You want to analyze them. You want to look at the charts and you want to go, okay, where is your, where are the companies that you sell into? Where are they going? And what does this tell about our business? So this is kind of the big idea where we said, okay, let's first start with like tapping into all these data sources.

[00:14:01] So we have a lot of, as you said, we have a lot of data sources that are just, you know, public data sources, such as, you know, consumer index and these things. We, of course, have, you know, company, company level data sources, stock prices, but also hiring signals and all these things we get from, for particular companies. And everything in between, we also track how an industry is going. So an industry is taken to a so-called NAICS code. And for these NAICS codes, you can also get a lot of data, like how a certain industry is developing also in a certain region.

[00:14:30] So we built this huge repository of data and first, and build forecasts on a company level. So for more than 4 million companies in the US, we can tell them, okay, the next 18 months, they will go in this direction or they will go in that direction. So that's kind of the thing that we built. It's about the first step. After we did this, we went out to pilot customers and worked with them. Hey, so this is what we have. We can predict businesses forward. How does this help you? What do we really need from us?

[00:15:01] And it turns out, again, we are a platform, right? Different pilot customers work with, have different use cases. And this is kind of where we gathered a lot of information and kind of steered and pivoted along our way to a kind of refined first use case that we heavily invest right now, which is basically prospecting engines. Well, one of the things I was going to bring up with you is a report that you recently released. And before you came on the podcast, I was skimming through it.

[00:15:29] And it introduces this new idea of quiet restructuring where AI-related workforce contraction may happen through hiring freezers, attribution, unfilled positions and automation, rather than those headline-grabbing mass layoffs that we keep seeing in our LinkedIn newsfeed. So what did you find in that data? And what might conventional employee statistics that we might be reading elsewhere, what are they essentially missing? Interesting. So interesting question.

[00:15:59] So first of all, when we go back, so there has been this big narrative that there is an AI apocalypse, right? Companies like AI is there and AI will take all the jobs of, you know, all the employees in the country and so on. And people will lose their job and everything will get automated. First of all, this is a narrative, right? So this is driven by a handful of companies.

[00:16:27] This is something we see in the data that the companies that are actually doing, you know, layoffs, that are announcing layoffs and they relate this to AI, it's only a handful of companies. It's a number that's less than, less than a hundred. So it's a kind of a narrative that's driven by a company. It's not like every, you know, out of the 4 million companies that we have in our database, we see that 50% doing this. No, no, no, no, no, no. It's like, it's like a, it's like a rounding error of the companies that are actually doing this.

[00:16:54] The second thing is that we actually see if we look at projections, like overall the company, we see that we, that, that the total, you know, workforce growth in the US, it's kind of flat. It's maybe a little bit increasing, but it's flat, but it's not, we are not on, on the brink of a cliff right now. So this is not what's happening. So if there's an, if there's an, if there's actually an AI related workforce reduction, it's going to be not a reduction total headcount.

[00:17:23] It's rather be a restructuring. So this is the hypothesis where we go into this report, where we say, okay, no, we don't see a decrease. We don't see that people are laying off employees in large cases. We rather see what we see in the data is that, for example, one thing is that we see that companies tend to adopt more and more technologies. We see there's a certain, there's a certain cluster of companies that are actually decreasing in headcount, but it's also not all the companies.

[00:17:51] We see that new roles are hired. We call them the transformation AI roles. These are roles like, you know, transformation manager, chief AI officer, these kinds of roles. We see them pop up for a lot of companies. And they especially also pop up for those companies that make these announcements that I just, just talked about. And we see that the, you know, that the internal structure, because we also have all this job and hiring data.

[00:18:16] We see that the structure within the companies, even if a company is flat, we see that the amount of leadership increases a little bit over time. This is something we see. We also see that companies tend to get a little bit older on average. So their workforce average age is getting a little bit older, which also tends to be a sign of, okay, we're not saying companies are, you know, straight up laying people off. But they're kind of, you know, they're kind of sitting in a backseat and watching right now.

[00:18:45] They may be slowed their hiring. They may be rethinking their strategy. They say, okay, let's push hiring to the next quarter or one quarter other. Let's observe a little bit what's happening before we make a final decision here. If we should buy something or hire something for that. So these are kind of the cried signals that we see in the data, where we say that's not a big layoff here. If there is something, there's a restructuring going on in these companies. And this was kind of our first step at the larger series of reports that we want to do.

[00:19:15] We want to, you know, I don't know if you know the parable of touching the elephant from different sides, right? So that's kind of one angle that we approach this whole topic of. But we also want to approach this topic from other different angles. And this is kind of what the data enables us to do. And I think that's very exciting. And we just want to give, you know, one more signal for the broader discussion, like shape the narrative a little bit with what we have at Carecast.

[00:19:41] And I'm curious, when you look at all those signals and everything that you've discovered there, what does it tell us about how AI is changing companies before the impact becomes publicly visible? And we see it on our news feeds. Anything else that the signals raise there? Anything that took you by surprise? I think the fact that it's only a couple of companies that are driving the narrative, this is something that took me by surprise.

[00:20:10] Because if you look at the news, you know, which is anecdotal, you think, okay, it's all over the place, which it is right, which it is not. We also did some, you know, I don't want to speak too much to this. But we also did some follow-through analysis that where we saw also like a big chunk of these companies that actually do these announcements, they don't follow through with it. So we see their hiring trend is a certain way. So they make this announcement. And when, sure, within the companies, the structure of the companies that changes.

[00:20:39] But if you look at like the top line of the headcount that they have, that trend doesn't change a lot after this announcement. So this is also something we see. Also, I must, you know, I must disclaim here that these announcements, they are, you know, mostly recent, they're mostly less than six months old.

[00:20:55] So to make a real, you know, significant follow-through analysis, we have to wait for at least another six months, which is due to the fact that the data that we see, it's always a little bit lagged, you know, people that are going to lay it off. So, for example, the hiring data that we get is kind of, you know, we're looking at, you know, what is out there and people don't immediately get laid off and then, I don't know, change their LinkedIn or change wherever they, you know, indicate their current role.

[00:21:24] So this data is always a little bit lagged. So in order to make a, you know, real conclusion, we got to wait a couple more months. But at least the early finding is that an announcement doesn't, you know, lead to an extra follow-through. And your report does explicitly state that it presents a hypothesis and that the model outputs are not facts and that the correlation does not establish causation.

[00:21:46] So as a data scientist, how do you communicate uncertainty and prevent people from treating a probability as a prediction of an inevitable layoff? I would imagine it could be quite a balance because there are a lot of journalists out there that would just want to take that worst-case scenario and just run with it, you know. That's a really good point. And I think that is a tension that, you know, that's in all companies. You always have, you know, the people that are, for example, like me, I'm the data scientist.

[00:22:16] I'm always like, okay, I actually only want to put the facts out there. I don't want to build a narrative out there. But of course, I also have marketing. They say, okay, but what could be hypothesized out of this data? And I'm always like, okay, yeah, it's correlation. We see that these things kind of go together, but we cannot say, okay, that's the reason here. That's the causality. So I'm always saying, okay, let's make sure we frame what we know as, you know, fact. But what we don't know, let's always caveat this.

[00:22:45] Okay, that's what the data suggests. It's an hypothesis, right? For example, we see that, you know, let's take this example. A company restructures internally, right? It could be just a random blip, right? We don't actually know that it's AI. We only see that it kind of aligns with, you know, the point in time that we are right now. So we assume companies are thinking about AI. So we assume that, you know, changes that we see in a company are largely driven by AI.

[00:23:11] But of course, as a data scientist, I must say, it could be anything. We don't know actually AI. So yeah, you're totally right. That's always a tension that we try to, you know, match out within the company 100%. So if predictive intelligence can give businesses, investors, policymakers and workers listening earlier visibility into organisational change, how do you think we should use this foresight responsibly?

[00:23:38] What decisions could we make differently if we identify those early signs of workforce transformation months before they appear in financial results or official employment data? They did say in Spider-Man with great power comes great responsibility. But what should we be doing here? That's a good question. I'm not sure if I have a good answer for you. I just want to be cautious. So what we should do is probably that we shouldn't do things premature before we know what the actual effects are.

[00:24:07] I must say, I personally in general are super optimistic. I don't think there will be, you know, big seismic shifts in the economy so fast that we are not able as a society to react to them in time. So I said earlier, so what we see is, you know, companies are, you know, reducing, decelerating, back hiring and so on. So I think we should learn from these companies that we say, okay, no, we don't want to do big changes right now. We want to see where the company goes.

[00:24:35] We want to probably not shackle them too much because what we always have seen in the past that with the new technologies, things change, but we don't know what's going to be to change. Right. Imagine you would have restricted something in the development in the Internet in the early 2000s because you would think the Internet would kill jobs. Right. But back in the day, you wouldn't even know what a data scientist job role is. Right. This job role wouldn't exist at the time. So that's always the thing. Right.

[00:25:03] Of course, some jobs, they will just go away. But these technologies, they will, you know, they will build a new, a whole new industry. Like there was back hundreds of new industry with hundreds of new job roles with people that are happy to work in these new job roles. So that's why, you know, just, just, just let it be and see how it, how it works out.

[00:25:25] And when you first started and came together and formed ClaireCast, your goal was to aspire to be the world's best predictor of a company headcount. Your mission has evolved since those early days. So what excites you now about the future? What you're working on? Anything that listeners can be aware of, of what's coming down the road? What are you, what excites you at the moment? So what excites me at the moment is actually that we see those shifts into agentic systems.

[00:25:53] And that's kind of a, that's kind of a buzzword, so to say. And I must totally admit, like, if you would talk to me about agents, like a year or one and a half year ago, I would say, okay, no, that's a buzzword. That's never happening. But especially in the last six months with these new frontier models from Anthropic, now also from OpenAI. Last week we saw something from Moonshot AI, the new Kami model.

[00:26:20] They've gotten so much better, especially in the six months, that I really can see that agentic systems where you just give a model a prompt. You let the model interview yourself to let the model understand, okay, what do you actually want to build here? And then it goes out and actually builds the thing for you that works super reliably right now. And this didn't work in the past.

[00:26:46] So how often was I sitting on my, you know, development environment? I was writing code. I was asking it just for one function or one single line. And I always had to check it. And then it didn't produce what I wanted. Now it can literally write the whole code base, not flawless. There's bugs in there. And it produces too much code, redundant code, dead code, whatever. But, like, it can get there.

[00:27:10] And this is something that I'm still digesting this right now. And I also know for myself, like, I'm not using these technologies, like, to the best of the extent that it should be possible to do this. Although I'm thinking that I'm already, you know, at the frontier of using these things. But when I'm looking at the socials and see what people are doing, they are building, I don't know, last weekend, I saw that someone led this new Kami model from Moonshot.

[00:27:40] He asked it to build Mac OS. And then he let it run over the weekend. And on Monday morning, he had Mac OS on his computer with, like, everything. There was a music app there. And then the music app was populated with songs. And then there was a message app. And he could, like, call his friends from this, you know, from this AI-coded tool. I'm not sure how much of this is, like, just PR and how much of this is true. But I think it points to something really, which is super exciting. Wow.

[00:28:10] Exciting times ahead. And we've referenced one of your reports a few times today. And obviously for people listening wanting to find out more information about that, about the work you're doing, keep up to speed with future announcements, connect with you or your team. Where's the best starting point for everything? So future announcements. So right now we are just released this report. We are right now thinking what's going to be the next report. So the first idea, of course, is to look at the other side of the coin.

[00:28:36] If we can learn something about companies where we think, right, not where we are certain, where we think that they are not quite restructurers. So companies that are not, you know, following this trend, where we don't see that they restructure internally, where we see that they are growing. They don't do any ANI announcement. We also see that they are not, you know, adopting a lot of technologies. So picking out now this bunch of companies and see what we can learn about them, maybe about the current moment. This is going to be one of the next reports that we want to publish.

[00:29:06] But then we want to, of course, also work on our first use case that I alluded to earlier on, which is going to be what we saw with a lot of the pilot customers that we worked on. They want to know, hey, you have all this information about where are companies going. Can you just give us information about leads? Can you, like, figure out who we should contact you to sell our product? And this is something that we tested out, which is working super well right now.

[00:29:30] Because what companies see right now is that there's right now a shift in how every company needs to go to market with AI. Because you probably have also gotten all these emails that are kind of personalized for you. But they're personalized very bad. It's like, hey, Neil, I liked your last podcast episode. Have you ever thought about, you know, I don't know, buying this certain soda product? I think it's like super.

[00:29:58] I mean, it's kind of prey and spray outbound marketing with AI slot these days. And it's not working. So no one is clicking these emails anymore. And companies see this. Their conversion rates drop massively right now. So what you need to do right now is to be much more strategic who you want to reach out to. You don't want to reach out to a thousand people that might fall in your ideal customer profile in your ICP.

[00:30:24] You want to reach out to exactly those 10 people where you are 100% sure they need your product. And when you reach out for them, you want to actually spend time and personalize the email and reach out to them in a, you know, in a sustainable way. This is possible with our data. We know exactly who your buyers probably are. We saw that this is working. We can lift a conversions rate for outreach a lot. So this is kind of the first use case that we now heavily invest within our larger platform.

[00:30:53] And this is what I'm excited about our product right now. Exciting times ahead. Well, I will have links to everything that you mentioned. I encourage people listening to go check you guys out. I'd love to stay in touch with you to see how your story evolves and get you back on in the near future. But more than anything, thank you for sharing your story today. Really appreciate your time. Lovely. And just want to mention it here. So if you want to check us out, we're at ClaireCast.com. Yeah. Thanks for having me, Neil. What a pleasure.

[00:31:20] One of the things I loved about Marvin's perspective is his willingness to keep uncertainty visible. And ClaireCast analyzes patterns across 4 million companies and produces 18-month forecasts. But he also reminds us that correlation does not establish causation. And announcements need time before their consequences can accurately be measured.

[00:31:45] And I think it's that discipline that matters, especially when the subject we're talking about is people's livelihoods. Because, yes, a headline can travel around the world before the underlying workforce data has even finished updating its LinkedIn profile. And the practical lesson here is to treat predictive intelligence as another signal for decision. Then combine it with context, human judgment and patience.

[00:32:13] And clercast's early findings suggest companies might be reorganizing roles and slowing down hiring without producing that broad collapse that is implied by some AI narratives. And when you are doomscrolling, remember, those bad stories get more clicks than good stories. So that essentially means that there are many journalists out there that will focus on just the bad stuff. All they care about, all many publications care about, is more clicks, more likes, more shares, rage bait.

[00:32:43] Yeah, I'm sure you've seen much of that out there. So a big thank you to Marvin for joining me on Tech Talks Daily. And you can find out more. So a big thank you to Marvin for joining me today. You can follow Marvin and Clarecast online for future reports and product updates. And remember, when you see the next headline predicting an AI jobs apocalypse or SaaSpocalypse, what evidence are you going to look at before you accept that story?

[00:33:10] And also pop over techtalksnetwork.com. That's where you'll find me hanging out. And I'll also be hanging out in your podcast feeds tomorrow morning. Hopefully I'll speak with you all then. Bye for now.